On 7 July 2016, an Airbus A320 (N333NW) being operated by Delta Airlines on a scheduled passenger flight from Minneapolis St. Paul MN to Rapid City SD as DAL 2845 was cleared for a visual approach to runway 14 at destination in day VMC but instead landed uneventfully on runway 13 at Ellsworth AFB which was located 6 miles northwest of the intended destination and was open at the time and had no conflicting traffic. Only when they were about to touch down did either pilot realise their error.
An Investigation was carried out by the NTSB. The 2 hour CVR and FDR were removed from the aircraft and their data were successfully downloaded.
It was noted that the 60 year-old Captain, who had been PF for the investigated flight had approximately 25,800 total flying hours which included 2,980 hours on type. He had been employed by Delta since it took over his previous employer Northwest Airlines in 2010 and had made just one arrival to and one departure from Rapid City in that time. The 51 year-old First Officer had 7,600 total flying hours including 2,324 hours on type. He had been employed by Delta since 2000 and had never flown to or from Rapid City. Both pilots were based in Salt Lake City.
It was established that the flight had proceeded uneventfully until reaching the vicinity of Rapid City. The crew had initially expected to be using runway 32 for landing and had briefed accordingly, but by the time they were about to begin positioning, the wind had changed and favoured runway 14. The Captain had, "due to his personal procedure […] prepared for the runway 14 approach as well (so) the change was not a significant factor". It was found that Delta charts for Rapid City included mention of its close proximity to Ellsworth Air Force Base to the northwest when on final approach and it was noted during the Investigation that the Ellsworth runway 13 extended centreline crosses the Rapid City 14 extended centreline at about 7½ miles prior to the 14 threshold. The approach briefing had included mention of this possible confusion.
As the aircraft descended through FL 235 for 17,000 feet, the crew made initial contact with Ellsworth APP and the flight was re-cleared to 5,300 feet and advised to expect a visual approach to runway 14. At this point the aircraft was approximately 45 nautical miles east of destination. A few minutes later, the flight was abeam destination and was instructed to fly a radar heading of 300° "for a downwind leg to the visual approach". The Ellsworth APP controller and the Rapid City TWR controller discussed on the landline that the aircraft "was high and fast for the visual approach" and whilst they were talking, the aircraft descended through about 12,000 feet. The field elevation of Rapid City was 3,200 feet and the nominal remaining flying distance of 15 to 18 miles indicated that the aircraft was "well above the typical 300 feet per mile descent". About the same time, the Captain noted that the speed was too high and then recognised that the way he had set up the FMS was not going to work as he had intended. He re-selected 'Open Descent' mode. A minute later he said "there's the airport," and called for gear down and flaps one. By this time the destination was to the south-southwest of the aircraft in an 8 o'clock relative position whilst Ellsworth was in a 10 o'clock relative position. This suggested to the Investigation that "it is likely the Captain was actually looking at the latter".
After delaying their base leg turn from downwind because they were still too high, the crew eventually accepted a radar heading onto a left turn as they descended through 5,800 feet about 12 miles from destination and advised ATC that they "had the field in sight". Their height was consistent with a descent to Rapid City runway 14 but would have been "somewhat steeper than normal angle" for a descent to Ellsworth runway 13 which was only about 5 miles away. Ellsworth APP then called with "cleared visual approach runway 14, use caution for Ellsworth Air Force Base located six miles northwest of Rapid City Regional". After acknowledging the clearance, the First Officer was recorded as saying to the Captain "you got the right one in sight?" to which the captain replied "I hope I do" and as the aircraft approached the intersection of the two extended centrelines, the First Officer asked if they could contact TWR and this was approved. At this point, the Captain switched off the AP and "directed the First Officer to clear the flight director display". Then, just after switching to TWR and about 1,300 feet above the elevation of the runways at both the destination and Ellsworth and only about 3 miles from the latter, the rate of descent steepened. The Captain stated that he had not noticed the PAPI lights for the runway he was descending towards (which would have been all white) and had remained "focused on the visual approach". With the aircraft 1.5 nm from Ellsworth runway 13 (and 8 miles from Rapid City runway 14), the Captain called "confirmed stable" with the aircraft descending at approximately 1,200 fpm and added "this is the most [expletive] approach I've made in a while".
As they approached the runway, the Captain began to retard the thrust levers to idle at which point both pilots reported having realised that they were landing at Ellsworth. The landing was reported to have been "uneventful" and they cleared the runway and notified Rapid City TWR of their location.
It was noted that "clear skies" had prevailed for the final approach and landing, and that sunset was approximately 4 minutes prior to the landing. It was further noted that the lighting for both runway 13 at Ellsworth and runway 14 at Rapid City had been on and fully functional and that although both had PAPIs, the runway at Ellsworth was twice as wide (90 metres) as that at Rapid City (45 metres) and significantly longer (over 4,000 metres compared to 2,650 metres). It was also noted that only Ellsworth was equipped with an approach lighting system which included sequenced flashing lights in addition to the high intensity runway lighting which was illuminating both runways. Finally, it was found that the airport data for Rapid City provided for Delta flight crew reference and carried on board contained an explicit warning that "Ellsworth AFB lies northwest of RAP on final approach for runway 14 - these airports have similar runway alignment and can be mistaken for one another". It was noted that a previous similar event, not investigated by the NTSB, had occurred in 2004 when a Northwest Airlines Airbus A319 also landed at Ellsworth instead of Rapid City.
Wrong Airport landings at other locations in the USA were noted as not uncommon and as including events involving civil air transport flights on a regular although not frequent basis - for example the visual approach to a landing at an airport 6 miles north of its intended destination, Branson MO, by a Southwest Airlines Boeing 737-700 in 2014.This Investigation led to Safety Recommendation 15-09 to the FAA to “amend air traffic control procedures so that controllers withhold landing clearance until the aircraft has passed all other airports that may be confused with the destination airport”. As of the publication of the Rapid City Investigation Final Report, this Recommendation remains classified as “Open - Unacceptable Response”. The Branson Investigation was noted during the Rapid City work as having also resulted in Safety Recommendation 15-10 to the FAA to "Modify the minimum safe altitude warning (MSAW) software to apply the MSAW parameters for the flight plan destination airport to touchdown, rather than automatically reassigning the flight to another airport based on an observed (and possibly incorrect) trajectory". The basis for this recommendation was noted as having been that “ATC radar data processing systems typically include MSAW functions that compare the aircraft's expected trajectory with its observed trajectory and alert controllers if the aircraft is in danger of collision with terrain or obstructions”. Since this “is accomplished by comparing the aircraft's altitude against a digital terrain model until it reaches the vicinity of the destination airport when the processing changes to compare the aircraft's observed trajectory against expected trajectories for landing aircraft”, it was considered that MSAW systems should be configured to detect when an aircraft is “unexpectedly descending to the ground away from the destination airport” and then generate an MSAW alert. In this respect, “a review of the Ellsworth radar data showed that as DAL2845 approached the Ellsworth area, the system applied MSAW rules for Ellsworth arrivals instead of Rapid City arrivals” so that no MSAW alert was generated. It was noted that the FAA response to this Recommendation was, at the time of completion of the Rapid City Investigation, classified as "Open – Acceptable Alternate Response” on the grounds that, although the FAA did not concur with the recommendation because “the current design of the MSAW tool does not contain any element to address it”, they did “plan to investigate other possible software solutions aside from MSAW” that would satisfy this Recommendation.
The Investigation found that the Probable Cause of the event was "the flight crew's misidentification of the desired landing runway due to excess altitude requiring an extended downwind, and failure to use all available navigation information”.
The Final Report of the Investigation was adopted and published on 26 May 2017. No further Safety Recommendations were made.